81P Single-cell and bulk RNA sequencing analysis of the pan-cancer immune microenvironment identifies a novel tumor-associated macrophage signature predicting immunotherapy response

Identifying robust biomarkers for predicting patient response remains a challenge in the field of tumor immunotherapy. Tumor-associated macrophages (TAMs), the most abundant cells in the innate immune system, significantly influence cancer prognosis and immunotherapy.We explored the potential association of a TAM molecular signature with prognosis and other relevant tumor characteristics to enhance prediction accuracy for patient response to immunotherapy. We conducted a comprehensive analysis of single-cell RNA sequencing data from two cancer datasets to dissect the molecular characteristics of TAMs. Next, we investigated potential relationships between TAMs.Sig and prognosis across 33 different types of cancer. We compared it with other previously reported six ICI response signatures and developed an immune response prediction model based on TAMs.Sig. Finally, we performed functional and subtyping analyses in non-small cell lung cancer patients using TAMs.Sig. We identified seven novel TAM prognostic genes, namely CRYAB, SHC1, CCL3L3, CREG1, PCDHGC3, CCND1, and XAGE1A. Subsequently, these genes were utilized to construct a TAMs.Sig. The presence of TAMs.Sig exhibits a negative correlation with the tumor immune response. The predictive model developed using KNN regression demonstrated superior diagnostic capability compared to other algorithms with an AUC value of 0.703. In patients with lung adenocarcinoma (LUAD), there exists a significant relationship between TAMs.Sig and both mDFS(p=0.034) as well as mOS(p=0.0061). Two different subtypes of LUAD were identified by important tumor macrophage prognostic genes and Cluster B exhibited higher dysregulation scores along with elevated CD274 expression. However,TMB was found to be higher in patients belonging to Cluster A compared to those in Cluster B. Our findings reveal that clustering patients based on TAMs.Sig predicts ICI outcomes with greater accuracy than other previously identified signatures across multiple types of cancer and is a promising approach for predicting ICI response in patients with lung adenocarcinoma.